PROJECT SUMMARY/ABSTRACT Ductal carcinoma in situ (DCIS) of the breast is a heterogeneous neoplasm that can give rise to invasive and metastatic breast cancer. The drivers of DCIS progression to invasive disease have been only partly understood and the Breast Cancer Initiating cell (BCIC) subpopulation in DCIS has not been well-defined due to lack of representative DCIS models. Still, it is known that the MCF-DCIS cell line that models transition of basal DCIS to invasive cancer in vivo, harbors bipotential progenitor BCIC. With this model we found that depletion of the nuclear coactivator Amplified In Breast cancer 1 (AIB1) reduced the number of CD44+/24- BCIC cell population in MCF- DCIS, leads to loss of the basal myoepithelial layer, to reduced tumor growth and to reduced progression to invasive disease. We propose that the properties of a BCIC subpopulation determines the ability of DCIS to progress to invasive disease. We have reported a novel method for propagating epithelial cells indefinitely using conditional reprogramming of cells (CRC) by defined culture conditions. In this proposal we wish to develop appropriate new CRC models of DCIS derived epithelial cell lines from human tumors that can be used in in vitro and in vivo studies to a) uncover the breast BCIC involvement in DCIS progression, b) discover pathways involved in DCIS to invasive progression. Aim 1: To establish a population of cells from human DCIS samples that can undergo DCIS to invasive transition in vivo. To predict if patient- derived DCIS CRC lines from different subtypes of breast cancer will likely be tumorigenic in vivo, we will first screen for their ability to form invasive or non-invasive spheroids in a 3D organotypic spheroid, matrigel /collagen matrix assay. Non-invasive and invasive spheroids will be collected from the matrix, expanded in the CRC system and tested separately, as well as in combination, for their ability to form DCIS lesions and progress to invasive cancer after injection into the mammary fat pad and by intra-nipple injection into the mammary duct of immune compromised mice. Aim 2: To define the characteristics of CRC cells that can form DCIS lesions in vivo. Patient derived cells grown as non invasive or invasive spheroids in the 3D organotypic assay, will be expanded in the CRC system, and analyzed for BCIC characteristics 1) by FACS analysis of cell surface markers 2) by tumorsphere assays in non- adherent conditions 3) by comparison of their gene expression and somatic mutation patterns. 4) Candidate molecular drivers and pathways for maintenance and progression of DCIS will be assessed in functional assays in 3D and in immune compromised mouse models, using knockdown and overexpression approaches Impact: The availability of new models representing the different subtypes of DCIS will enable us to better define BCIC populations, and potential pathways and biomarkers that predict invasive potential of DCIS & aiding design of rational therapeutic targets to prevent DCIS progression.